JohnBradley

iTrader: (30)

So after much thinking about how to explain what true Speed Density is vs what we do here for instance, how AEM's are often setup, etc. I came up with the following idea, just explain what everything is and remove doubt.

MAF - if you dont know this one its a volume sensor. Thats it. Ours have IAT and a baro comp to make it more accurate but it simply measures volume entering the engine. If there are boost leaks its inaccurate, doesnt like VTA (letting metered air leave without pulling fuel), all the stuff we know.

Alpha-N - TPS x RPM. That is all that is with no VE compensation (i.e. boost). If I tell it that at 100% throttle and 6000 rpm it does it at all boost levels independent of air volume.

Speed density - estimates air mass by pressure vs temp. Then applies this to a 3D map. The "3D" we already have, its Load (pressure or volume) x RPM with a set value in each cell. Every point can be fine tuned so this is really Alpha N with compensation. Any VE changes needs a retune, doesnt matter the system. It only knows pressure and rpm not what the true mass of the air is (it has a trim table and most get ignored above certain levels).

THEN THERE IS WHAT I DO (and many others at this point).

I use the fuel temp sensor to mimic an IAT since its already in the car. The IAT reads gross temp not average. It can heat soak, be subject to a meth jet to close, all sorts of little issues. The actual trim table in the Evo ECU really isnt setup for IAT trim anyway. It has an algorithm that allows for minor alterations to fueling since its set for a MAF, and then trims timing vs air temp.

Fuel temp is more or less constant. The fuel heatsoaks going thru the rail hits the tank and then is sent forward again. It takes 15 minutes to get full heat into the fuel system (ask anyone with an A1000). It is a rough average of engine bay temp which is what the MAF does factory. I have done exhaustive testing and see no more than 5-6* variation from fuel temp as a source to what the MAF would read in the corner of the engine bay.

Now for the term everyone has been waiting for, Boost Comp. This is what MAFs do, most SD setups (AEM, Motec, Autronic, Vipec, etc.) do for fueling. It assumes VE is more or less constant until you hit backpressure in the turbine housing, head flow, cam size, and VE starts to drop off again.

I set the fueling at 16psi and in our maps this is 180% load or so. The values are constant out to 38psi. Hyper Boost Compensation is the proper name. Retuning is as easy as altering either the max boost portion of the MAPVE table to allow more or less fuel vs pressure, and in some cases the tip in portion of the MAPVE (121kpa column).

There is no such thing as true speed density. I havent seen any AEM ever do this or any other standalone. GM is the closest to True SD and its because tada they use a MAF as well to judge volume vs pressure.

Do I do it right? I dunno. Does it work well when its setup, I build a good map, and then work that map on each and every application? Yes. Are there advantages and disadvantages? Of course, before I had the MAF comp vs Coolant the cars DID NOT love life cold. The IXs were better but not perfect, the VIIIs didnt like me.

....to be continued....

badev0

iTrader: (74)

Good read!

seventh_rx

iTrader: (2)

Thank you for posting this!

JohnBradley

iTrader: (30)

There is also Alpha-N with Boost Comp. This is more or less what we do EXCEPT IT HAS NO TEMP input. It is solely rpm x TPS x pressure. Since we aren't doing fueling by TPS vs RPM at any point other than accel enrichment it is not accurate to call the SD we normally do on Evos Alpha N or N-Alpha or Alpha-Omega or whatever. Alpha N is different and is normally something reserved for ITB cars or quads (single cylinders) that have really non specific fueling and just need fuel vs tps x rpm.

06EvoIXmr07

iTrader: (22)

I never knew you didnt have to use an ait sensor,do you have to do anything special to the rom to use the fuel temp sensor?Im using the tephra prepatched Ix rom.

l2r99gst

iTrader: (2)

I give you an A for effort, but some of your thinking is off. The temp sensor in the MAF isn't there to measure engine bay temp. It's there to measure intake air at the maf sensor.

As you correctly stated, the Mitsu Karman MAF is a volume air sensor. The baro and temp are needed to correct to mass airflow. That's why the baro and temp sensor are directly in the MAF, taking the measurments where the volume air flow is being measured.

As you know, there are many 'loads' in our ECU. The load that is used for most of the fuel and ignition 3D maps is the baro+temp compensated load, meaning it's the mass airflow. So, our maps aren't pressure or volume vs RPM, it's actual mass airflow vs RPM. Of course, there are certain conditions that the ECU uses where it uses an uncompensated load where only volumetric airflow is used (such as using only baro comp load for the ignition maps over IAT of 77F, to disallow too much advance), but in general, the mass airflow is used so that it can calculate the mass of fuel.

It's basic PV=nRT. In an engine, volumetric airflow is:

Airflow (CFM) = PR[RPM*V.E.*Cid/3456]
PR=Pressure ratio=(boost in psi+atmos(psi))/atmos(psi)
RPM = RPM of engine
V.E. = volumetric efficiency at RPM being measured
Cid=cubic inch displacement= 122 for our 2.0L engines

To get mass airflow, you have to multiply that volumetric airflow by P/RT:

Airflow(lb/min)=(Airflow(CFM)*P/RT)*29(g/mol air)
P=atmostpheric pressure in PSI
R=ideal gas constant=10.7316 ft3·psi· °R-1·lb-mol-1
T=temperature in R (F +460)


In terms of SD for the Evo ECU, the IAT sensor needs to know the true air temp after the intercooler. That's the only way it's going to know the true mass airflow. Using the fuel temp sensor isn't telling it the correct info. Of course, you can tune around it, with the 3D maps and the VE maps, but you are simply correcting the mistaken mass airflow that is being reported.

211Ratsbud

iTrader: (1)

So for instance, would running the MAF IAT>>> with a gm IAT sensor in the cold piping be about as efficient as one could get in terms of achieving accurate fueling?

l2r99gst

iTrader: (2)

Yes, that's why the IAT needs to be post IC. Either in UICP or IM. The only issue is that heatsoaking of the IAT (for example after engine shutdown and starting a warm engine later) would cause inaccurate fueling. I think DSMLink's SD implementation on the DSM ECUs uses different parameters for startup to alleviate this issue. This can potentially be patched to the EVO ECU SD as well. Aside from that, the current EVO SD implementation with GM IAT post IC should have very accurate mass airflow calculations.

But, in my opinion, using the fuel temp sensor or just some engine bay temp sensor isn't a good idea, especially for the most important parts, such as WOT, where the IATs can climb rapidly. Your fueling will be off by quite a bit. Of course, it would be off on the rich side, which is safer.

I would imagine Aaron doesn't see much variation because he has (1) good intercooling, so temps don't rise too much during a pull, (2) open filter, so intake temps are starting close to engine bay temps, (3) an efficient setup all around making a lot of power, where he isn't in the throttle for huge periods of time, where intake temps can dramatically change. Even if so, it can all be tuned around, but that is more of a band-aid kind of fix than I would like, personally.

Boosted Tuning

iTrader: (12)

I disagree...

A SD system or a system with VE tables, will need to be retuned for any VE changes, as the SD system cant compensate for VE changes.

A MAF based system will compensate for the change in VE, but will need to be retuned for max performance.

2000max

iTrader: (3)

While I see the point in your post. I think there is something to be said about the repeatability and consistency of using the fuel temp as a replacement. I can compensate for it "being off" and have still consistent fueling since it does not act in an irratic manner despite not being a true iat temp.

I hope this makes sense despite speaking in generalities.

KevinD

iTrader: (56)

definately agree with you l2r99gst. using the fuel temp as IAT is an awful idea for road race cars where true heatsoak of the intercooler with throw the tune way off. up in WA it might not be an issue, but down here in TX heatsoak is a huge issue. for drag cars or cars that don't see prolonged high turbo outlet temps, its not going to be a problem like it is for a track car.

l2r99gst

iTrader: (2)

Yes, I completely see your point and I agree that if you are in an environment and have a setup that lends itself to be tuned using the fuel temp or some other arbitrary temp sensor, then it will be fine.

You just have to know that and be aware that you aren't calculating or measuring any true mass airflow and that any deviation in that temp that you aren't measuring will throw off the tune.

Boosted Tuning

iTrader: (12)

When you say that fuel temp mimics air temp, I thought you were talking about post IC air temp, and not just the MAF temp, which is pre IC.

So have you compared fuel temp to air temp using a real AIT sensor and NOT the MAF temp sensor?

KevinD

iTrader: (56)

its definately MAF temps, because post IC temps can vary drastically on what intercooler you use, what turbo, how much boost, and of course the air coming into the turbo. and also varys over time depending on airflow through the intercooler, and heatsoak.

Boosted Tuning

iTrader: (12)

Exactly.

I made that post because JB has been posting that fuel temp mimics air temp, and I thought he was basing that statement off comparing the fuel temps to air temps taken with an actually AIT sensor, not the MAF. Now that he states he making this comparison off MAF temps, Id like to know if he's compared fuel temps to actual air temp taken with a AIT sensor.